Pain and Pain Management
Understanding pain | assessing pain | managing pain | general principles | specific pain syndromes | medications | addiction | opioids/narcotics | nonopioids | routes of administration | other measures (updated 1/11/2005)
Many people, doctors, nurses, and patients alike, live in the past where pain is concerned. Bound by dysfunctional and erroneous traditions such as, "pain should be expected" or "we need to save the strong medicine until the pain is really bad" or "respiratory depression is a common complication of narcotic use in patients with cancer," they sentence their patients to unnecessary suffering.
Pain, often the most feared aspect of cancer, can be managed with oral medications in as many as 90% of patients with cancer pain, and by alternative means in the other 10%. Unfortunately, factors such as inadequate assessments, lack of knowledge among doctors and nurses, problems in communications among staff, and patient, family, and staff myths and expectations about pain often result in poorly managed pain, especially in the case of chronic non-malignant pain. Pain may also be poorly managed because some patients do not report pain out of fear that pain means the illness is worsening or that painful tests may result from the complaint. Older people in particular tend to not report or to deny the existence of pain.
|
To adequately manage moderate to severe pain it is necessary to understand how to:
|
There are two general types of pain and they differ significantly. First, acute pain is caused by tissue damage and it diminishes as the tissue heals. Acute pain lasts for hours to days, and is often accompanied by physical signs such as rapid heartbeat, sweating, pallor, and inability to sleep. Examples of acute pain include pain from a broken arm or surgery.
Chronic pain, on the other hand, is pain that lasts or recurs for months or even years. Chronic pain does not often decrease of its own accord and after the first weeks or months, is not accompanied by physical signs. Thus a person with chronic pain does not have an increased heart rate, sweating, pallor; and he or she is able to sleep to some extent.
Unrelieved chronic pain has a negative effect on the psychological, social, and spiritual life of the person in pain. Psychologically, there is a clear link between chronic pain and anxiety, depression, and suicide ideation and attempts. Socially, people with chronic pain tend to be more isolated because of irritability and inability to tolerate interactions with others. Spiritually, chronic pain tends to result in people feeling hopeless; that life (and their unrelieved suffering) is meaningless, and that they are apart from God.
The physiological classification of pain is of great - and sometimes unrecognized - importance in how pain is treated. Physiologically, pain is classified as nociceptive (somatic or visceral), neuropathic, or mixed:
Somatic (or nociceptive) pain is caused by damage or stimulation to the limbs and/or body wall, including skin and deep tissue such as muscle and bone. Bone metastasis and surgery are common causes of somatic pain in patients with cancer. Somatic pain is usually constant and fairly localized; and is described as aching, throbbing, gnawing, etc.
Visceral (or nociceptive) pain is caused by distention, compression, inflammation, or other stimulation of the internal organs. Visceral pain is usually constant and generalized; and is described as deep, squeezing, generalized, and constant; but may also be lancinating (sharp); and may also be referred to the skin.
Neuropathic (or deafferentation) pain is caused by damage to or stimulation of nerves, most commonly as a result of neoplasm, but also from radiation injury as well as processes unrelated to cancer, e.g., diabetic neuropathy. Neuropathic pain may be constant or intermittent, and sometimes shock-like; and is usually described as sharp, stabbing, burning, or as an unpleasant sensation.
In general, neuropathic pain is the most severe and difficult to manage of the three basic classes of pain. Some patients have a combination of these classes of pain. It is not unusual for a patient with advanced cancer to have more than one site of pain (e.g., back and abdominal) and more than one type of pain (e.g., somatic and neuropathic). There are some specific pain syndromes, the most troublesome, but not all of which, are discussed below. Readers wishing to pursue this further are referred to current texts such as deVita et al, Cancer: Principles and Practice of Oncology, The American Pain Society, The Journal of Pain and Symptom Management, and other current sources of information.
| "I'm not at all sure that the pain suffered by a previously well person for a brief period of days after a surgical operation is comparable to long-continued suffering which affects both mind and body... Nor do I believe that any kind of measurements of pain intensity or drug comparisons can be relied upon... because the base-line is constantly shifting" (William K. Livingston, 1955). |
To knowledgeable clinicians, the patient's perceptions of the pain are, by far, the most important part of assessment. One of the most respected American experts on pain says, simply, "Believe the patient's complaint of pain." Effective assessments of pain must be consistent and systematic. The parameters of the initial and periodic pain assessment are listed below. Assessments of the most basic parameters (location, severity, and quality) should be carried out every hour or two until pain is brought under control. If pain is stable and well-managed, assessments every few days are usually adequate. Pain assessment parameters include:
Location(s) and radiation. Where is the greatest pain and to where does it extend? Most pain assessment tools include a front and back line drawing of a human figure. Patients mark the areas that are painful and clinicians can also mark areas of known metastasis, surgical sites, and other potential causes or influences on pain.
Severity. Scales are often used to measure severity. The particular kind of scale may be of less importance than for everyone involved (physicians, nurses, families, others) to use the same one, as in the final analysis, scales are always arbitrary and subjective. Scales may use numbers such as 1-10; words such as none, mild, moderate, severe, and overwhelming; or a visual analogue scale (VAS), which uses a straight 10 cm. line with one end representing no pain and the other the worst possible pain. Colored VAS scales are also in use.
Quality. The patient describes the pain, using words such as dull, aching, sharp, shooting, burning, numb but feels bad, etc. Understanding quality gives insight into the specific type of pain.
Influencing factors (other than medications) and patterns. What increases or decreases the pain? When is the pain most and least severe? Is the pain constant or does it come and go? What feelings or emotional or spiritual factors affect the pain?
Other physical symptoms, particularly those that are new or increasing. Rapid changes in pain, consciousness, nausea, ability to walk, balance, and ability to hold objects are especially important.
Impact of pain on daily life, including (1) activity, (2) mood, (3) enjoyment of life (4) sleep, (5) sociability, and (6) any other relevant aspect of life.
Meaning of the pain to the patient and family. Some people believe pain is just part of sickness or is punishment from God and should thus be borne in silence.
Pain history. Did the pain occur suddenly or develop gradually; after a particular event (such as changes in medications, a coughing spell, a fall, etc.)? What medications and/or treatments have controlled the pain, and what failed?
Effects of current medication(s). Baseline data includes (1) name of medication(s), (2) effects, (3) length of effects, (4) concerns of patient and family about the medication(s), (5) history of medication(s) taken for pain, (6) any over-the-counter, homeopathic, or herbal remedies used for pain.
An acronym used by some in the assessment of pain is OLDCART:
In addition, laboratory or other diagnostic tests may be used to help clinicians understand pain and other problems. The decision whether to use diagnostic tests should be carefully made and not just an automatic response to problems. Clinicians should work to balance the need for testing with the patient's quality of life. There may not be a clear-cut answer to whether tests are truly needed for the patient's benefit.
Opioids, especially morphine given orally in the controlled release (CRM) form (e.g., MS Contin, Oramorph SR) are the primary means of managing moderate to severe pain in advanced cancer. Prescribed, taken, and titrated (adjusted) correctly, oral opioids effectively control pain in as many as 90% of patients with cancer pain. Managing pain is not, however, a matter of just taking one's medicine; it is necessary to follow the established principles of cancer pain management.
Controlled release morphine was a major breakthrough in pain management. Prior to its availability, we were having to give patients medications at least every four hours; and in some cases, large numbers of tablets. That's six times over a 24 hour period, so there was always at least one time when we would have to awaken the patient. For awhile some clinicians were trying methadone because it has a longer half-life and duration of action than morphine. Unfortunately, methadone tends to accumulate in body tissue and can eventually cause difficulties, especially for older patients. Now with controlled release morphine, the patient takes pain medications 2-3 times a day. Its really much better.
General Principles of Cancer Pain Management
| Summary of Effective Means of Pain Management: Take adequate amounts of medication at regular intervals so that pain is prevented. It is usually possible to manage pain with oral controlled release morphine + another medication such as aspirin, antidepressant, or steroid. Additional medications should be on-hand in case breakthrough pain occurs. |
The following are general guidelines to management of the specific types of pain. In all cases, the best option usually is to remove or decrease the cause of pain, e.g., shrink a tumor with radiation. Addressing the cause of the pain is not possible in many cases, especially in advanced disease. In most cases, a combination of medications of different classes is used (i.e., an opioid and a non-steroidal anti-inflammatory drug; not two different opioids). The types of pain are managed as follows:
Opioids, primarily morphine, are the chief means of managing cancer pain, with morphine taken orally effective in the great majority of patients. Although new medications and new delivery systems are constantly being developed, the central issue in pain management is not new technology, but the skillful use of what is known to work.
Notes on Addiction (The Problem That Isn't - Where Cancer Pain is Concerned)
Addiction (psychological and physical dependence on opioids, including compulsive drug-seeking behavior) is feared by many patients and professionals. However, despite scare stories, media hysteria, and professional ignorance, the fact is that the risk of addiction is insignificant in patients who take opioids for pain and who do not have a history of drug abuse.
Tolerance, the need for increasing amounts or frequency of medication to manage pain, does develop in some patients taking opioids for cancer pain, but increasing the amount of medication and/or decreasing the interval at which it is given is usually all that is needed to counter the tolerance. Difficulty in initially controlling pain is not due to tolerance, but to inadequate dosing. Where strong opioids and cancer pain are concerned, there is no point at which medications are no longer effective because of tolerance. Approximately 50% of patients with cancer pain require increasing doses of opioids, usually with an initial rapid, then slower increase in amount. About 30% of patients take approximately the same dose and 20% decrease the dose over time. Tolerance can also be addressed by switching opioids and/or route of administration.
The scenario that results in the greatest desire for opioid medications and the greatest degree of "clock-watching" is when medications are not strong enough, or are given in insufficient amounts, or at too lengthy intervals. Waiting to give medications results in more pain and less of an effect of the medication on the pain.
| Patients might ask themselves this question, "If I were not having this pain, would I be out looking for drugs?" If the answer is "no," the chances of addiction are insignificant. Properly used, pain medications allow the patient to control the pain rather than be controlled by the pain. |
Strong Opioid (Narcotic) Analgesics
The opioid analgesics most effective and most commonly used for moderate to severe pain are morphine, hydromorphone (Dilaudid), and fentanyl (Duragesic). Oxycodone was previously considered effective only for mild-moderate pain, but is now considered by most to be a strong opioid analgesic. Morphine, hydromorphone, and oxycodone are available in controlled release (CR) formulations. Heroin has been used in England, but is not superior to morphine. Meperidine (Demerol) is a commonly used for postoperative pain but is a poor choice for regular use in cancer pain because of side effects that occur with regular use.
Side effects occasionally limit the use of opioids such as morphine or Dilaudid; but far more often, limits in opioid use are based on misunderstanding of the medications themselves or their side effects and how to address them. The most common side effects of opioids include:
Mild Opioid Analgesics
Mild opioids include codeine (including Tylenol with codeine) and propoxyphene (Darvon, Darvocet). Although many clinicians view these medications as effective in mild-moderate cancer pain, they are, except for oxycodone, effective primarily in mild pain; and then when given in combination with another medication such as aspirin or Tylenol. Oxycodone is more effective than codeine or propoxyphene for mild or moderate pain. Propoxyphene is ineffective for anything other than mild pain in many patients, and has tendency to accumulate in the body with chronic use. Most mild opioids are sold in a fixed-dose combination with other medications such as aspirin or Tylenol. The problem with this form of medication is that while the dose of opioid (e.g., oxycodone or codeine) may be increased without adverse effects, the dose of the other medication should not be significantly increased because of adverse side effects. The side effects of the mild opioids are similar to those of the strong opioids.
Non-opioid analgesics include non-steroidal anti-inflammatory drugs (NSAIDs) such as aspirin, indomethacin (Indocin), ibuprofen (Motrin, Advil), and similar medications. Acetaminophen (Tylenol) has analgesic, but not antiinflammatory effects. NSAIDs are the analgesics of first choice for mild pain and, given with a strong opioid, are effective in managing bone pain from metastasis, even when the pain is severe. The Cox-2 inhibitors such as celecoxib (Celebrex) are very effective NSAID medications, with fewer adverse gastrointestinal effects than other NSAIDs. In 2004 one of the COX-2 inhibitors (Vioxx) was pulled from the market because of concerns about cardiaovascular effects.
Although Tylenol has fewer adverse side effects than aspirin, it should be used with caution in patients with alcoholism and/or liver disease (including liver metastasis). Aspirin and other NSAIDs have adverse effects on the gastrointestinal and renal systems, as well as the liver, especially with regular use. All the mild non-opioid analgesics have a ceiling effect in which analgesia does not increase at higher doses and toxicity occurs.
Mixed Agonist/antagonist and Partial Agonist Opioids
Opioid agonist/antagonists include pentazocine (Talwin), butorphanol (Stadol), and nalbuphine (Nubain). Buprenorphine (Buprenex) is a partial agonist and is effective in moderate to severe pain. Patients with chronic cancer pain should avoid these medications because they produce less analgesia than the strong opioids or oxycodone; have greater side effects than opioids; and also precipitate withdrawal in patients who take opioids.
Adjuvant Medications
Adjuvant medications are used to treat pain that does not completely respond to opioids, increase the analgesic effects of opioids, decrease opioid side effects, and manage associated symptoms. Adjuvant medications include the previously discussed NSAIDS, corticosteroids, tricyclic antidepressants, anticonvulsants, phenothiazines, and antianxiety medications.
Corticosteroids: Dexamethasone (Decadron), prednisolone (Cortalone, Delta-Cortef), and prednisone are commonly used corticosteroids in patients with cancer. Corticosteroids decrease inflammation and thus decrease pain; and also improve appetite, strength, and sense of well-being. Specific indications for corticosteroids include pain from bony metastasis, nerve involvement, and joint pain resulting from cancer. Specific problems (that may include pain) in which corticosteroids are used include increased intracranial pressure (headache), spinal cord compression (back pain), and superior vena cava syndrome (chest pain). Rapid withdrawal of corticosteroids increases pain and sleep disturbances may occur. It is necessary then to gradually decrease the amounts taken and to not take the medication at night. Because of numerous and cumulative side effects, corticosteroids should seldom be taken indefinitely.
Antidepressants: The most commonly used antidepressants in patients with cancer are amitriptyline (Elavil), doxepin (Sinequan), and imipramine (Tofranil). Although sometimes used for depression and/or aiding sleep, these antidepressants also have analgesic effects, especially in neuropathic pain from postherpatic neuralgia, nerve involvement in cancer; and in multiple sclerosis or diabetes. Tricyclic antidepressants are most often used when neuropathic pain is continuous and are used in combination with an opioid. Antidepressants for pain and/or sleep are taken lower doses than for depression; and their effects are usually experienced in one to two days rather than the two-four weeks required in depression. Side effects may be transitory, and include dry mouth, blurred vision, and constipation. Selective serotonin reuptake inhibiting (SSRI) antidepressants such as Prozac also have an important place in treatment of depression, but (except for Paxil) are not used in pain management.
Anticonvulsants: Commonly used anticonvulsants include carbamazepine (Tegratol), phenytoin (Dilantin), and clonazepam (Klonopin). These are used in combination with an opioid for neuropathic pain, especially when the pain is intermittent; but also for post-herpetic neuralgia and when pain is continuous. Regular measurement of blood levels of anticonvulsants, complete blood counts, and assessment for infection are necessary when using anticonvulsants.
Phenothiazines: The only phenothiazine with proven analgesic properties is methotrimeprazine (Levoprome), which is used primarily in patients with bowel obstruction. The most frequently used phenothiazine in patients with cancer is prochlorperazine (Compazine), used for nausea. Phenothiazines such as Thorazine are used for extreme anxiety, psychomotor agitation, insomnia, bladder or rectal pain, and ureteral spasm.
Anti-anxiety Medications: Although often prescribed in terminal illness, anti-anxiety medications do not have analgesic properties. In many cases, the anxiety for which anti-anxiety medications are prescribed responds better to either improved pain management or management of depression (since anxiety is often part of depression). In other cases, decreasing anxiety through the use of anti-anxiety medications helps indirectly to decrease the pain.
Oral medications, given in adequate amounts, at regular and correct intervals, and usually in combination with adjuvant medication(s) are nearly always effective in controlling pain. Oral medications are also easier to give, less expensive than injections, and far less expensive than high tech means such as patient-controlled analgesia.
Intervals between doses are based on the duration of action of the medication. The interval between doses is a commonly overestimated by clinicians who are unskilled in managing pain. The peak drug effect for most opioids is reached in 1.5 - 2 hours. Controlled release opioids take longer to reach maximum effect. According to the American Pain Society, if relief from pain is not achieved with morphine or Dilaudid in that time (1.5-2 hours), and side effects are not troublesome, the patient can usually safely take a second dose.
| Equianalgesia: When changing from one route to another (for example, from injections or intravenous to oral) it is imperative that equianalgesic (producing equal analgesia) doses be given. It takes about three times as much morphine taken orally, for example, to produce the same analgesia as morphine taken intramuscularly or intravenously. Thus a patient whose pain is well-managed with 30 mg of morphine by injection every four hours will need about 90 mg orally every four hours; or if switching from 30 mg of morphine by injection every four hours to continuous release oral morphine with a 12 hour duration of action, the same patient will need about 180 mg of continuous release morphine every 12 hours. The equianalgesic ratio varies according to the specific drug used. |
Because it furnishes a lengthier duration of action and is thus more convenient, controlled release (CR) morphine (MS Contin, Avinza, or Kadian) has replaced morphine sulfate or Dilaudid as the drug of choice for pain management. MS Contin's duration of action is 12 hours, but eight hour intervals are occasionally required. Avinza's duration of action is 24 hours and Kadian's duration of action is 12-24 hours.
Opioid Equianalgesic Doses (non-CR formulations)
| Medication | Parenteral dose (mg) | Oral dose (mg) | Half-life (hours) | Duration of action (approximate) |
| Morphine | 10 |
30 |
2-3 |
3-4 |
| Hydromorphone | 2-3 |
7.5 |
2-3 |
2-4 |
| Oxycodone | 15 |
20-30 |
2-3 |
2-5 |
| Fentanyl* | 0.1 |
- |
2-4 |
0.5-4 |
| Meperidine | 75 |
300 |
2-3 |
2-4 |
* Equianalgesic doses (fentanyl and morphine) usually given may be too conservative and result in increased pain when switching from morphine to fentanyl. A 2:1 (mg/day of oral morphine: microgram/hour of transdermal fentanyl) may be more effective (see references: Skaer, 2004).
Medications for breakthrough pain or special circumstances should always be available. Because of the fast action, injections are preferable. When injections are not feasible, non-controlled release oral or rectal medications are acceptable. Other effective routes include:
Radiation therapy
Radiation therapy may be used to shrink tumors that are causing pain. The use of radiation is limited by the responsiveness of the tumor to radiation, previous radiation in the same area, and by the need to transport the patient to a facility to receive therapy. Several weeks may pass before radiation has an effect and thus other means of controlling pain should also be used. There are instances in which radiation causes edema and a temporary increase in symptoms. Radiation therapy clearly has a place in palliative care.
Chemotherapy
Depending on a variety of factors, chemotherapy affects or prevents pain by reducing tumor size and inhibiting certain pathological processes.
Anesthetic Measures
Anesthetic measures or nerve blocks are most often used for severe pain refractory to opioid and other more common therapies. Phenol or alcohol injections are given by an anesthesiologist at specific "trigger points" or in specific nerves. Permanent blocks may be used if temporary blocks are effective.
Surgery
Neurosurgical approaches to pain are used in specific circumstances when other measures are unsuccessful.
Psychological and Alternative Means of Controlling Pain
Psychological and alternative measures for controlling pain in patients with significant cancer pain are useful as adjuncts to opioids and other proven means of pain control. Meditation, distraction, visualization, relaxation, etc. are all helpful, but are not the final answer in pain control. Social support, counseling, and hypnosis are also helpful as adjuncts to therapy, but again, are not the mainstay. Biofeedback and transcutaneous nerve stimulation are helpful to some patients, but are also best used as part of an overall plan of therapy.
Some Common and WRONG Ways to Manage Moderate to Severe Cancer Pain
References
Chen ML, Chang HK. (2004). Physical symptom profiles of depressed and nondepressed patients with cancer. Palliative Medicine, 18, 712-718.
Cherny, N.I. (2002). Cancer pain: Principles of assessment and syndromes. In A.M. Berger, R.K. Portenoy, & D.E. Weissman (Eds.), Principles & practice of palliative care & supportive oncology (2nd. ed., pp. 3-52). Philadelphia: Lippincott Williams & Wilkins.
Cleeland, C.S. (1990). Assessment of pain in cancer: Measurement issues. In K.M. Foley, J.J. Bonica, and V. Ventafridda (Eds.). Proceedings of the Second International Congress on Cancer Pain (pp. 47-55). New York: Raven Press.
Fernandez, E. (2002). Anxiety, depression, and anger in pain. Dallas: Advanced Psychological Resources.
Gatchel, R.J. (2004). Biofeedback as an adjunctive treatment modality in pain management. APS Bulletin, 14 (American Pain Society). Retrieved November 30, 2004 from http://www.ampainsoc.org/pub/bulletin/jul04/clin1.htm
Gordon, D.B., Dahl, J., Phillips, P. Frandsen, J., Cowley, C., Foster, R.L., Fine, P.G., Miaskowski, C., Fishman, S. & Finley, R.S. (2004). The Use of ‘As-Needed’ Range Orders for Opioid Analgesics in the Management of Acute Pain: A Consensus Statement of the American Society for Pain Management Nursing and the American Pain Society. APS Bulletin, 14 (American Pain Society). Retrieved November 30, 2004 from http://www.ampainsoc.org/pub/bulletin/jul04/consensus1.htm
Hall PW, Christie PM, Currow D. (2005). Cannabinoids and cancer: Causation, remediation, and palliation. Lancet Oncology, 6, 35-42.
Jaeckle KA. (2004). Neurological manifestations of neoplastic and radiation-induced plexopathies. Seminars in Neurology, 24, 385-393.
Kemp, C.E. (1999). Terminal Illness: A Guide to Nursing Care (2nd ed.). Philadelphia: Lippincott-Williams & Wilkins.
Kessler, R.S. (2004). Is it time to propose a continuum of self-regulation treatments for the control of pain? Evidence for effectiveness of hypnosis and relaxation. APS Bulletin, 14 (American Pain Society). Retrieved November 30, 2004 from http://www.ampainsoc.org/pub/bulletin/mar04/clin2.htm
Livingston, W.K. (1953). Quoted in M. Meldrum (1999). Travels in the east: A Livingston story. APS Bulletin, 9 (American Pain Society). Retrieved December 12, 2004 from http://www.ampainsoc.org/pub/bulletin/mar99/history.htm
Massie, M.J. and Holland, J.C. (1992). The cancer patient with pain: Psychiatric complications and their management. Journal of Pain and Symptom Management, 7, 99-109.
Mercadante, S.G. (1998). When oral morphine fails in cancer pain: The role of alternative routes. American Journal of Hospice and Palliative Care, 15, 333-342.
Paice, J.A., Toy, C., & Shott, S. (1998). Barriers to cancer pain relief: Fear of tolerance and addiction. Journal of Pain and Symptom Management, 16, 1-9.
Paice, J.A. (2004). Pain. In C.H. Yarbro, M.H. Frogge, & M. Goodman (Eds.), Cancer symptom management (3rd ed., pp. 77-93). Boston: Jones and Bartlett.
Perry, S. and Heidrich, G. (1982). Management of pain during debridement: A survey of U.S. burn units. Pain, 13, 267-280.
Portenoy, R.K. (2000). Current pharmacotherapy of chronic pain. Journal of Pain and Symptom Management, 19(1 - Supplement), S16-S20.
Porter, J. and Jick, H. (1980). Addiction rare in patients treated with narcotics. New England Journal of Medicine, 302, 123.
Sawaki, Y., Parker, R.K., and White, P.F. (1992). Patient and nurse evaluation of patient-controlled analgesia delivery systems for postoperative pain management. Journal of Pain and Symptom Management, 7, 443-453.
Sorge, J. & Sittl, R.(2004). Transdermal buprenorphine in the treatment of chronic pain: Results of a phase III, multicenter, randomized, double-blind, placebo-controlled study. Clinical Therapeutics, 26, 1808-1820.
SUPPORT Investigators. (1995). A controlled trial to improve care for seriously ill hospitalized patients. Journal of the American Medical Association, 274, 1591-1598.
Sullivan, M. & Ferrell, B. (2005). Ethical challenges in the management of chronic nonmalignant pain: Negotiating through the cloud of doubt. Journal of Pain, 6, 2-9.
Vogl, D., Rosenfeld, B., Breitbart, W., Thaler, H., Passik, S., McDonald, M., & Portenoy, R.K. (1999). Symptom prevalence, characteristics, and distress in AIDS outpatients. Journal of Pain and Symptom Management, 18, 253-262.
Skaer, T.L.. (2004). Practice guidelines for transdermal opioids in malignant pain. Drugs, 64, 2629-2638.